Balanced graph partitioning is a critical step for many large-scale distributed computations with relational data. As graph datasets have grown in size and density, a range of highly-scalable balanced partitioning algorithms have appeared to meet varied demands across different domains. As the starting point for the present work, we observe that two recently introduced families of iterative partitioners---those based on restreaming and those based on balanced label propagation (including Facebook's Social Hash Partitioner)---can be viewed through a common modular framework of design decisions. With the help of this modular perspective, we find that a key combination of design decisions leads to a novel family of algorithms with notably better empirical performance than any existing highly-scalable algorithm on a broad range of real-world graphs. The resulting prioritized restreaming algorithms employ a constraint management strategy based on multiplicative weights, borrowed from the restreaming literature, while adopting notions of priority from balanced label propagation to optimize the ordering of the streaming process. Our experimental results consider a range of stream orders, where a dynamic ordering based on what we call ambivalence is broadly the most performative in terms of the cut quality of the resulting balanced partitions, with a static ordering based on degree being nearly as good.
more »
« less
Partitioning Communication Streams Into Graph Snapshots
We present EASEE (Edge Advertisements into Snapshots using Evolving Expectations) for partitioning streaming communication data into static graph snapshots. Given streaming communication events (A talks to B), EASEE identifies when events suffice for a static graph (a snapshot ). EASEE uses combinatorial statistical models to adaptively find when a snapshot is stable, while watching for significant data shifts – indicating a new snapshot should begin. If snapshots are not found carefully, they poorly represent the underlying data – and downstream graph analytics fail: We show a community detection example. We demonstrate EASEE's strengths against several real-world datasets, and its accuracy against known-answer synthetic datasets. Synthetic datasets' results show that (1) EASEE finds known-answer data shifts very quickly; and (2) ignoring these shifts drastically affects analytics on resulting snapshots. We show that previous work misses these shifts. Further, we evaluate EASEE against seven real-world datasets (330 K to 2.5B events), and find snapshot-over-time behaviors missed by previous works. Finally, we show that the resulting snapshots' measured properties (e.g., graph density) are altered by how snapshots are identified from the communication event stream. In particular, EASEE's snapshots do not generally “densify” over time, contradicting previous influential results that used simpler partitioning methods.
more »
« less
- Award ID(s):
- 1916084
- NSF-PAR ID:
- 10388359
- Date Published:
- Journal Name:
- IEEE Transactions on Network Science and Engineering
- ISSN:
- 2334-329X
- Page Range / eLocation ID:
- 1 to 18
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
There has been a growing interest in the graph-streaming setting where a continuous stream of graph updates is mixed with graph queries. In principle, purely-functional trees are an ideal fit for this setting as they enable safe parallelism, lightweight snapshots, and strict serializability for queries. However, directly using them for graph processing leads to significant space overhead and poor cache locality. This paper presents C-trees, a compressed purely-functional search tree data structure that significantly improves on the space usage and locality of purely-functional trees. We design theoretically-efficient and practical algorithms for performing batch updates to C-trees, and also show that we can store massive dynamic real-world graphs using only a few bytes per edge, thereby achieving space usage close to that of the best static graph processing frameworks. To study the applicability of our data structure, we designed Aspen, a graph-streaming framework that extends the interface of Ligra with operations for updating graphs. We show that Aspen is faster than two state-of-the-art graph-streaming systems, Stinger and LLAMA, while requiring less memory, and is competitive in performance with the state-of-the-art static graph frameworks, Galois, GAP, and Ligra+. With Aspen, we are able to efficiently process the largest publicly-available graph with over two hundred billion edges in the graph-streaming setting using a single commodity multicore server with 1TB of memory.more » « less
-
We give an $\widetilde{O}(\sqrt{n})$-space single-pass 0.483-approximation streaming algorithm for estimating the maximum directed cut size (Max-DICUT) in a directed graph on n vertices. This improves over an $O(\log n)$-space $4 / 9 < 0.45$ approximation algorithm due to Chou, Golovnev, and Velusamy (FOCS 2020), which was known to be optimal for $o(\sqrt{n})$-space algorithms. Max-DICUT is a special case of a constraint satisfaction problem (CSP). In this broader context, we give the first CSP for which algorithms with $\widetilde{O}(\sqrt{n})$- space can provably outperform $o(\sqrt{n})$- space algorithms. The key technical contribution of our work is development of the notions of a first-order snapshot of a (directed) graph and of estimates of such snapshots. These snapshots can be used to simulate certain (non-streaming) Max-DICUT algorithms, including the “oblivious” algorithms introduced by Feige and Jozeph (Algorithmica, 2015), who showed that one such algorithm Previous work of the authors (SODA 2023) studied the restricted case of bounded-degree graphs, and observed that in this setting, it is straightforward to estimate the snapshot with $\ell_{1}$ errors and this suffices to simulate oblivious algorithms. But for unbounded-degree graphs, even defining an achievable and sufficient notion of estimation is subtle. We describe a new notion of snapshot estimation and prove its sufficiency using careful smoothing techniques, and then develop an algorithm which sketches such an estimate via a delicate process of intertwined vertex- and edge-subsampling. Prior to our work, the only streaming algorithms for any CSP on general instances were based on generalizations of the $O(\log n)$-space algorithm for Max-DICUT, and can roughly be characterized as based on “zeroth” order snapshots. Our work thus opens the possibility of a new class of algorithms for approximating CSPs by demonstrating that more sophisticated snapshots can outperform cruder ones in the case of Max-DICUT.more » « less
-
Recently, considerable research attention has been paid to graph embedding, a popular approach to construct representations of vertices in latent space. Due to the curse of dimensionality and sparsity in graphical datasets, this approach has become indispensable for machine learning tasks over large networks. The majority of the existing literature has considered this technique under the assumption that the network is static. However, networks in many applications, including social networks, collaboration networks, and recommender systems, nodes, and edges accrue to a growing network as streaming. A small number of very recent results have addressed the problem of embedding for dynamic networks. However, they either rely on knowledge of vertex attributes, su er high-time complexity or need to be re-trained without closed-form expression. Thus the approach of adapting the existing methods designed for static networks or dynamic networks to the streaming environment faces non-trivial technical challenges. These challenges motivate developing new approaches to the problems of streaming graph embedding. In this paper, we propose a new framework that is able to generate latent representations for new vertices with high e ciency and low complexity under speci ed iteration rounds. We formulate a constrained optimiza- tion problem for the modi cation of the representation resulting from a stream arrival. We show this problem has no closed-form solution and instead develop an online approximation solution. Our solution follows three steps: (1) identify vertices a ected by newly arrived ones, (2) generating latent features for new vertices, and (3) updating the latent features of the most a ected vertices. The new representations are guaranteed to be feasible in the original constrained optimization problem. Meanwhile, the solution only brings about a small change to existing representations and only slightly changes the value of the objective function. Multi-class clas- si cation and clustering on ve real-world networks demonstrate that our model can e ciently update vertex representations and simultaneously achieve comparable or even better performance compared with model retraining.more » « less
-
null (Ed.)The proliferation of GPS-enabled devices has led to the development of numerous location-based services. These services need to process massive amounts of streamed spatial data in real-time. The current scale of spatial data cannot be handled using centralized systems. This has led to the development of distributed spatial streaming systems. Existing systems are using static spatial partitioning to distribute the workload. In contrast, the real-time streamed spatial data follows non-uniform spatial distributions that are continuously changing over time. Distributed spatial streaming systems need to react to the changes in the distribution of spatial data and queries. This article introduces SWARM, a lightweight adaptivity protocol that continuously monitors the data and query workloads across the distributed processes of the spatial data streaming system and redistributes and rebalances the workloads as soon as performance bottlenecks get detected. SWARM is able to handle multiple query-execution and data-persistence models. A distributed streaming system can directly use SWARM to adaptively rebalance the system’s workload among its machines with minimal changes to the original code of the underlying spatial application. Extensive experimental evaluation using real and synthetic datasets illustrate that, on average, SWARM achieves 2 improvement in throughput over a static grid partitioning that is determined based on observing a limited history of the data and query workloads. Moreover, SWARM reduces execution latency on average 4 compared with the other technique.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Journal Article:
- https://doi.org/10.1109/TNSE.2022.3223614
